Accurate Prediction of a Quantitative Trait Using the Genes Controlling the Trait for Gene-Based Breeding in Cotton

Area of Science:

• Plant Genetics
• Quantitative Trait Prediction
• Agricultural Biotechnology

Background:

• Accurate phenotype prediction is crucial for advancing plant research and breeding programs.
• Quantitative traits, like cotton fiber length, present challenges for precise prediction.
• Existing genomic selection methods using random markers have limitations in prediction accuracy.

Purpose of the Study:

• To develop a highly accurate method for predicting cotton fiber length using specific genes.
• To identify key cotton fiber length (GFL) genes essential for accurate phenotype prediction.
• To assess the efficiency of gene-based prediction models compared to traditional genomic selection.

Main Methods:

• Utilized 474 cotton fiber length (GFL) genes and nine prediction models.
• Employed Single Nucleotide Polymorphisms/Insertions-Deletions (SNPs/InDels) and gene expression data for prediction.
• Identified key predictive genes through analysis of GFL gene contributions.

Main Results:

• Achieved a prediction accuracy of r = 0.83 using SNPs/InDels from 226 GFL genes or expression of all 474 GFL genes.
• This represents an 116% improvement over previous prediction accuracies for cotton fiber length.
• Identified 125 key GFL genes that enable prediction accuracy comparable to using all 474 GFL genes.
• Demonstrated high consistency of prediction accuracies across different environments and generations.
• Determined that a training population of 100-120 plants is sufficient for accurate model training.

Conclusions:

• Genes controlling a quantitative trait can accurately predict its phenotype.
• Gene-based prediction significantly enhances the accuracy and efficiency of phenotype prediction in plants.
• This approach holds promise for advancing gene-based breeding strategies in cotton and other species.